Group members : Harsh, Rutvik, Jay¶

Project Name : Clustering Algoritham with python¶

1)Objective¶

  • The main objective of this research work is to analyze and represent the various factors that have contributed to the Evolution of credit card transaction over time in Graphical Format and perform a comparative study between various factors.
  • Find the relationship between the credit limit with purchases and payments.
  • Find the clusters in which data lies for the credit card data.
  • Find the different types of clustering techniques such as hierarchical clustering, K-means, and sklearn. cluster, and kNN function.
  • For this project, Visa company is Stakeholders because they can easily understand the data based on our project and make important decisions easily.

Import required libraries such as pandas, NumPy, Matpltlib, and seaborn.

In [2]:
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.colors as mcolors
import seaborn as sns
import plotly.express as px

Read the file using read_csv command

In [3]:
credit_data = pd.read_csv("CC GENERAL.csv")
In [4]:
credit_data.head(5)
Out[4]:
CUST_ID BALANCE BALANCE_FREQUENCY PURCHASES ONEOFF_PURCHASES INSTALLMENTS_PURCHASES CASH_ADVANCE PURCHASES_FREQUENCY ONEOFF_PURCHASES_FREQUENCY PURCHASES_INSTALLMENTS_FREQUENCY CASH_ADVANCE_FREQUENCY CASH_ADVANCE_TRX PURCHASES_TRX CREDIT_LIMIT PAYMENTS MINIMUM_PAYMENTS PRC_FULL_PAYMENT TENURE
0 C10001 40.900749 0.818182 95.40 0.00 95.4 0.000000 0.166667 0.000000 0.083333 0.000000 0 2 1000.0 201.802084 139.509787 0.000000 12
1 C10002 3202.467416 0.909091 0.00 0.00 0.0 6442.945483 0.000000 0.000000 0.000000 0.250000 4 0 7000.0 4103.032597 1072.340217 0.222222 12
2 C10003 2495.148862 1.000000 773.17 773.17 0.0 0.000000 1.000000 1.000000 0.000000 0.000000 0 12 7500.0 622.066742 627.284787 0.000000 12
3 C10004 1666.670542 0.636364 1499.00 1499.00 0.0 205.788017 0.083333 0.083333 0.000000 0.083333 1 1 7500.0 0.000000 NaN 0.000000 12
4 C10005 817.714335 1.000000 16.00 16.00 0.0 0.000000 0.083333 0.083333 0.000000 0.000000 0 1 1200.0 678.334763 244.791237 0.000000 12

2) Data Exploration¶

Find the rows and column of the dataframe.

Rows = 8950

Column = 18

In [5]:
credit_data.shape
Out[5]:
(8950, 18)

Find all statistical parameters such as mean, median, max, min, quartiles, and count for understanding data.

In [6]:
credit_data.describe().T
Out[6]:
count mean std min 25% 50% 75% max
BALANCE 8950.0 1564.474828 2081.531879 0.000000 128.281915 873.385231 2054.140036 19043.13856
BALANCE_FREQUENCY 8950.0 0.877271 0.236904 0.000000 0.888889 1.000000 1.000000 1.00000
PURCHASES 8950.0 1003.204834 2136.634782 0.000000 39.635000 361.280000 1110.130000 49039.57000
ONEOFF_PURCHASES 8950.0 592.437371 1659.887917 0.000000 0.000000 38.000000 577.405000 40761.25000
INSTALLMENTS_PURCHASES 8950.0 411.067645 904.338115 0.000000 0.000000 89.000000 468.637500 22500.00000
CASH_ADVANCE 8950.0 978.871112 2097.163877 0.000000 0.000000 0.000000 1113.821139 47137.21176
PURCHASES_FREQUENCY 8950.0 0.490351 0.401371 0.000000 0.083333 0.500000 0.916667 1.00000
ONEOFF_PURCHASES_FREQUENCY 8950.0 0.202458 0.298336 0.000000 0.000000 0.083333 0.300000 1.00000
PURCHASES_INSTALLMENTS_FREQUENCY 8950.0 0.364437 0.397448 0.000000 0.000000 0.166667 0.750000 1.00000
CASH_ADVANCE_FREQUENCY 8950.0 0.135144 0.200121 0.000000 0.000000 0.000000 0.222222 1.50000
CASH_ADVANCE_TRX 8950.0 3.248827 6.824647 0.000000 0.000000 0.000000 4.000000 123.00000
PURCHASES_TRX 8950.0 14.709832 24.857649 0.000000 1.000000 7.000000 17.000000 358.00000
CREDIT_LIMIT 8949.0 4494.449450 3638.815725 50.000000 1600.000000 3000.000000 6500.000000 30000.00000
PAYMENTS 8950.0 1733.143852 2895.063757 0.000000 383.276166 856.901546 1901.134317 50721.48336
MINIMUM_PAYMENTS 8637.0 864.206542 2372.446607 0.019163 169.123707 312.343947 825.485459 76406.20752
PRC_FULL_PAYMENT 8950.0 0.153715 0.292499 0.000000 0.000000 0.000000 0.142857 1.00000
TENURE 8950.0 11.517318 1.338331 6.000000 12.000000 12.000000 12.000000 12.00000

Check the types of the different parameters.

Found that every column has the proper type.

In [7]:
credit_data.dtypes
Out[7]:
CUST_ID                              object
BALANCE                             float64
BALANCE_FREQUENCY                   float64
PURCHASES                           float64
ONEOFF_PURCHASES                    float64
INSTALLMENTS_PURCHASES              float64
CASH_ADVANCE                        float64
PURCHASES_FREQUENCY                 float64
ONEOFF_PURCHASES_FREQUENCY          float64
PURCHASES_INSTALLMENTS_FREQUENCY    float64
CASH_ADVANCE_FREQUENCY              float64
CASH_ADVANCE_TRX                      int64
PURCHASES_TRX                         int64
CREDIT_LIMIT                        float64
PAYMENTS                            float64
MINIMUM_PAYMENTS                    float64
PRC_FULL_PAYMENT                    float64
TENURE                                int64
dtype: object

Print the name of all columns

CUST ID: Id of Credit Card holder

BALANCE: Balance amount left in their account

BALANCE FREQUENCY: How frequently the Balance is updated, score between 0 and 1 (1 = frequently updated, 0 = not frequently updated)

PURCHASES: Amount of purchases

ONE-OFF PURCHASES: Maximum purchase amount done in one-go

INSTALLMENTS PURCHASES: Amount of purchase done in installment

CASH ADVANCE: Cash in advance given by the user

PURCHASES FREQUENCY: How frequently the Purchases are being made score between 0 and 1 (1 = frequently purchased, 0 = not frequently purchased)

ONE-OFF PURCHASES FREQUENCY: How frequently Purchases are happening in one go (1 = frequently purchased, 0 = not frequently purchased)

PURCHASES INSTALLMENTS FREQUENCY: How frequently purchases in installments are being done (1 = frequently done, 0 = not frequently done)

CASH ADVANCE FREQUENCY: How frequently is the cash in advance being paid

CASH ADVANCE TRX: Number of Transactions made with Cash in Advance

PURCHASES TRX: Number of purchase transactions made

CREDIT LIMIT: Limit of Credit Card

PAYMENTS: Amount of Payment done by the user

MINIMUM_PAYMENTS: Minimum amount of payments made by the user

PRC FULL PAYMENT: Percent of full payment paid by the user

TENURE: Tenure of credit card service

In [8]:
credit_data.columns
Out[8]:
Index(['CUST_ID', 'BALANCE', 'BALANCE_FREQUENCY', 'PURCHASES',
       'ONEOFF_PURCHASES', 'INSTALLMENTS_PURCHASES', 'CASH_ADVANCE',
       'PURCHASES_FREQUENCY', 'ONEOFF_PURCHASES_FREQUENCY',
       'PURCHASES_INSTALLMENTS_FREQUENCY', 'CASH_ADVANCE_FREQUENCY',
       'CASH_ADVANCE_TRX', 'PURCHASES_TRX', 'CREDIT_LIMIT', 'PAYMENTS',
       'MINIMUM_PAYMENTS', 'PRC_FULL_PAYMENT', 'TENURE'],
      dtype='object')

3) Data Cleaning¶

Find the data which has null value by using the isna function.

And found that minimum_payments has 313 null value.

In [9]:
pd.isna(credit_data).sum()
Out[9]:
CUST_ID                               0
BALANCE                               0
BALANCE_FREQUENCY                     0
PURCHASES                             0
ONEOFF_PURCHASES                      0
INSTALLMENTS_PURCHASES                0
CASH_ADVANCE                          0
PURCHASES_FREQUENCY                   0
ONEOFF_PURCHASES_FREQUENCY            0
PURCHASES_INSTALLMENTS_FREQUENCY      0
CASH_ADVANCE_FREQUENCY                0
CASH_ADVANCE_TRX                      0
PURCHASES_TRX                         0
CREDIT_LIMIT                          1
PAYMENTS                              0
MINIMUM_PAYMENTS                    313
PRC_FULL_PAYMENT                      0
TENURE                                0
dtype: int64

Remove null value by passing 0 value.

In [10]:
credit_data=credit_data.fillna(0)
In [11]:
credit_data.head()
Out[11]:
CUST_ID BALANCE BALANCE_FREQUENCY PURCHASES ONEOFF_PURCHASES INSTALLMENTS_PURCHASES CASH_ADVANCE PURCHASES_FREQUENCY ONEOFF_PURCHASES_FREQUENCY PURCHASES_INSTALLMENTS_FREQUENCY CASH_ADVANCE_FREQUENCY CASH_ADVANCE_TRX PURCHASES_TRX CREDIT_LIMIT PAYMENTS MINIMUM_PAYMENTS PRC_FULL_PAYMENT TENURE
0 C10001 40.900749 0.818182 95.40 0.00 95.4 0.000000 0.166667 0.000000 0.083333 0.000000 0 2 1000.0 201.802084 139.509787 0.000000 12
1 C10002 3202.467416 0.909091 0.00 0.00 0.0 6442.945483 0.000000 0.000000 0.000000 0.250000 4 0 7000.0 4103.032597 1072.340217 0.222222 12
2 C10003 2495.148862 1.000000 773.17 773.17 0.0 0.000000 1.000000 1.000000 0.000000 0.000000 0 12 7500.0 622.066742 627.284787 0.000000 12
3 C10004 1666.670542 0.636364 1499.00 1499.00 0.0 205.788017 0.083333 0.083333 0.000000 0.083333 1 1 7500.0 0.000000 0.000000 0.000000 12
4 C10005 817.714335 1.000000 16.00 16.00 0.0 0.000000 0.083333 0.083333 0.000000 0.000000 0 1 1200.0 678.334763 244.791237 0.000000 12
In [12]:
pd.isna(credit_data).sum()
Out[12]:
CUST_ID                             0
BALANCE                             0
BALANCE_FREQUENCY                   0
PURCHASES                           0
ONEOFF_PURCHASES                    0
INSTALLMENTS_PURCHASES              0
CASH_ADVANCE                        0
PURCHASES_FREQUENCY                 0
ONEOFF_PURCHASES_FREQUENCY          0
PURCHASES_INSTALLMENTS_FREQUENCY    0
CASH_ADVANCE_FREQUENCY              0
CASH_ADVANCE_TRX                    0
PURCHASES_TRX                       0
CREDIT_LIMIT                        0
PAYMENTS                            0
MINIMUM_PAYMENTS                    0
PRC_FULL_PAYMENT                    0
TENURE                              0
dtype: int64

drop the customer id and prc full payment means percentage of the full payment in the data

In [13]:
credit_data.drop
Out[13]:
<bound method DataFrame.drop of      CUST_ID      BALANCE  BALANCE_FREQUENCY  PURCHASES  ONEOFF_PURCHASES  \
0     C10001    40.900749           0.818182      95.40              0.00   
1     C10002  3202.467416           0.909091       0.00              0.00   
2     C10003  2495.148862           1.000000     773.17            773.17   
3     C10004  1666.670542           0.636364    1499.00           1499.00   
4     C10005   817.714335           1.000000      16.00             16.00   
...      ...          ...                ...        ...               ...   
8945  C19186    28.493517           1.000000     291.12              0.00   
8946  C19187    19.183215           1.000000     300.00              0.00   
8947  C19188    23.398673           0.833333     144.40              0.00   
8948  C19189    13.457564           0.833333       0.00              0.00   
8949  C19190   372.708075           0.666667    1093.25           1093.25   

      INSTALLMENTS_PURCHASES  CASH_ADVANCE  PURCHASES_FREQUENCY  \
0                      95.40      0.000000             0.166667   
1                       0.00   6442.945483             0.000000   
2                       0.00      0.000000             1.000000   
3                       0.00    205.788017             0.083333   
4                       0.00      0.000000             0.083333   
...                      ...           ...                  ...   
8945                  291.12      0.000000             1.000000   
8946                  300.00      0.000000             1.000000   
8947                  144.40      0.000000             0.833333   
8948                    0.00     36.558778             0.000000   
8949                    0.00    127.040008             0.666667   

      ONEOFF_PURCHASES_FREQUENCY  PURCHASES_INSTALLMENTS_FREQUENCY  \
0                       0.000000                          0.083333   
1                       0.000000                          0.000000   
2                       1.000000                          0.000000   
3                       0.083333                          0.000000   
4                       0.083333                          0.000000   
...                          ...                               ...   
8945                    0.000000                          0.833333   
8946                    0.000000                          0.833333   
8947                    0.000000                          0.666667   
8948                    0.000000                          0.000000   
8949                    0.666667                          0.000000   

      CASH_ADVANCE_FREQUENCY  CASH_ADVANCE_TRX  PURCHASES_TRX  CREDIT_LIMIT  \
0                   0.000000                 0              2        1000.0   
1                   0.250000                 4              0        7000.0   
2                   0.000000                 0             12        7500.0   
3                   0.083333                 1              1        7500.0   
4                   0.000000                 0              1        1200.0   
...                      ...               ...            ...           ...   
8945                0.000000                 0              6        1000.0   
8946                0.000000                 0              6        1000.0   
8947                0.000000                 0              5        1000.0   
8948                0.166667                 2              0         500.0   
8949                0.333333                 2             23        1200.0   

         PAYMENTS  MINIMUM_PAYMENTS  PRC_FULL_PAYMENT  TENURE  
0      201.802084        139.509787          0.000000      12  
1     4103.032597       1072.340217          0.222222      12  
2      622.066742        627.284787          0.000000      12  
3        0.000000          0.000000          0.000000      12  
4      678.334763        244.791237          0.000000      12  
...           ...               ...               ...     ...  
8945   325.594462         48.886365          0.500000       6  
8946   275.861322          0.000000          0.000000       6  
8947    81.270775         82.418369          0.250000       6  
8948    52.549959         55.755628          0.250000       6  
8949    63.165404         88.288956          0.000000       6  

[8950 rows x 18 columns]>
In [14]:
credit_data.drop(['TENURE','PRC_FULL_PAYMENT'], axis=1,inplace = True)
In [15]:
credit_data.head()
Out[15]:
CUST_ID BALANCE BALANCE_FREQUENCY PURCHASES ONEOFF_PURCHASES INSTALLMENTS_PURCHASES CASH_ADVANCE PURCHASES_FREQUENCY ONEOFF_PURCHASES_FREQUENCY PURCHASES_INSTALLMENTS_FREQUENCY CASH_ADVANCE_FREQUENCY CASH_ADVANCE_TRX PURCHASES_TRX CREDIT_LIMIT PAYMENTS MINIMUM_PAYMENTS
0 C10001 40.900749 0.818182 95.40 0.00 95.4 0.000000 0.166667 0.000000 0.083333 0.000000 0 2 1000.0 201.802084 139.509787
1 C10002 3202.467416 0.909091 0.00 0.00 0.0 6442.945483 0.000000 0.000000 0.000000 0.250000 4 0 7000.0 4103.032597 1072.340217
2 C10003 2495.148862 1.000000 773.17 773.17 0.0 0.000000 1.000000 1.000000 0.000000 0.000000 0 12 7500.0 622.066742 627.284787
3 C10004 1666.670542 0.636364 1499.00 1499.00 0.0 205.788017 0.083333 0.083333 0.000000 0.083333 1 1 7500.0 0.000000 0.000000
4 C10005 817.714335 1.000000 16.00 16.00 0.0 0.000000 0.083333 0.083333 0.000000 0.000000 0 1 1200.0 678.334763 244.791237

4) Adaptation¶

Add a column about whether that customer is a regular customer or HNI.

If the customer's bank balance is more than 2000, the customer is HNI otherwise, the customer is regular.

In [16]:
credit_data["High_net_indi"] = 0
In [17]:
for i in range(0,8950):
  if credit_data.iloc[i,2] >= 2000:
    credit_data.iloc[i,16] = "HNI"
  else:
    credit_data.iloc[i,16] = "Regular customer"
In [18]:
credit_data[credit_data.iloc[:,2] <= 2000].head(3)
Out[18]:
CUST_ID BALANCE BALANCE_FREQUENCY PURCHASES ONEOFF_PURCHASES INSTALLMENTS_PURCHASES CASH_ADVANCE PURCHASES_FREQUENCY ONEOFF_PURCHASES_FREQUENCY PURCHASES_INSTALLMENTS_FREQUENCY CASH_ADVANCE_FREQUENCY CASH_ADVANCE_TRX PURCHASES_TRX CREDIT_LIMIT PAYMENTS MINIMUM_PAYMENTS High_net_indi
0 C10001 40.900749 0.818182 95.40 0.00 95.4 0.000000 0.166667 0.0 0.083333 0.00 0 2 1000.0 201.802084 139.509787 Regular customer
1 C10002 3202.467416 0.909091 0.00 0.00 0.0 6442.945483 0.000000 0.0 0.000000 0.25 4 0 7000.0 4103.032597 1072.340217 Regular customer
2 C10003 2495.148862 1.000000 773.17 773.17 0.0 0.000000 1.000000 1.0 0.000000 0.00 0 12 7500.0 622.066742 627.284787 Regular customer

Add new column about discount.

Provide a discount to a customer when the purchase is more than 5000, so the discount is 500.

If the purchase is more than 10000, so a discount is 1000.

If the purchase is more than 20000 so, a discount is 2000.

In [19]:
credit_data['Discount'] = 0
In [20]:
for i in range(0,8950):
  if credit_data.iloc[i,4] >= 20000:
    discount = 2000
    credit_data.iloc[i,17] = 2000
  elif credit_data.iloc[i,4] >= 10000:
    discount = 1000
    credit_data.iloc[i,17] = 1000
  elif credit_data.iloc[i,4] >= 5000:
    discount = 500
    credit_data.iloc[i,17] = 500
In [21]:
credit_data[credit_data.iloc[:,4] >= 20000].head(3)
Out[21]:
CUST_ID BALANCE BALANCE_FREQUENCY PURCHASES ONEOFF_PURCHASES INSTALLMENTS_PURCHASES CASH_ADVANCE PURCHASES_FREQUENCY ONEOFF_PURCHASES_FREQUENCY PURCHASES_INSTALLMENTS_FREQUENCY CASH_ADVANCE_FREQUENCY CASH_ADVANCE_TRX PURCHASES_TRX CREDIT_LIMIT PAYMENTS MINIMUM_PAYMENTS High_net_indi Discount
273 C10284 5131.319584 1.0 32539.78 26547.43 5992.35 0.0 1.000000 1.000000 1.000000 0.0 0 98 11500.0 26652.34432 1971.129435 Regular customer 2000
501 C10523 13479.288210 1.0 41050.40 40624.06 426.34 0.0 0.833333 0.666667 0.416667 0.0 0 157 17000.0 36066.75068 15914.484620 Regular customer 2000
507 C10529 2643.343414 1.0 26402.39 22257.39 4145.00 0.0 1.000000 1.000000 0.333333 0.0 0 114 16500.0 24529.27662 534.032339 Regular customer 2000
In [22]:
credit_data.head(10)
Out[22]:
CUST_ID BALANCE BALANCE_FREQUENCY PURCHASES ONEOFF_PURCHASES INSTALLMENTS_PURCHASES CASH_ADVANCE PURCHASES_FREQUENCY ONEOFF_PURCHASES_FREQUENCY PURCHASES_INSTALLMENTS_FREQUENCY CASH_ADVANCE_FREQUENCY CASH_ADVANCE_TRX PURCHASES_TRX CREDIT_LIMIT PAYMENTS MINIMUM_PAYMENTS High_net_indi Discount
0 C10001 40.900749 0.818182 95.40 0.00 95.40 0.000000 0.166667 0.000000 0.083333 0.000000 0 2 1000.0 201.802084 139.509787 Regular customer 0
1 C10002 3202.467416 0.909091 0.00 0.00 0.00 6442.945483 0.000000 0.000000 0.000000 0.250000 4 0 7000.0 4103.032597 1072.340217 Regular customer 0
2 C10003 2495.148862 1.000000 773.17 773.17 0.00 0.000000 1.000000 1.000000 0.000000 0.000000 0 12 7500.0 622.066742 627.284787 Regular customer 0
3 C10004 1666.670542 0.636364 1499.00 1499.00 0.00 205.788017 0.083333 0.083333 0.000000 0.083333 1 1 7500.0 0.000000 0.000000 Regular customer 0
4 C10005 817.714335 1.000000 16.00 16.00 0.00 0.000000 0.083333 0.083333 0.000000 0.000000 0 1 1200.0 678.334763 244.791237 Regular customer 0
5 C10006 1809.828751 1.000000 1333.28 0.00 1333.28 0.000000 0.666667 0.000000 0.583333 0.000000 0 8 1800.0 1400.057770 2407.246035 Regular customer 0
6 C10007 627.260806 1.000000 7091.01 6402.63 688.38 0.000000 1.000000 1.000000 1.000000 0.000000 0 64 13500.0 6354.314328 198.065894 Regular customer 500
7 C10008 1823.652743 1.000000 436.20 0.00 436.20 0.000000 1.000000 0.000000 1.000000 0.000000 0 12 2300.0 679.065082 532.033990 Regular customer 0
8 C10009 1014.926473 1.000000 861.49 661.49 200.00 0.000000 0.333333 0.083333 0.250000 0.000000 0 5 7000.0 688.278568 311.963409 Regular customer 0
9 C10010 152.225975 0.545455 1281.60 1281.60 0.00 0.000000 0.166667 0.166667 0.000000 0.000000 0 3 11000.0 1164.770591 100.302262 Regular customer 0

5)Descriptive Visualization:¶

plot the histogram of the purchases, purchases frequency, balnace, oneoff purchases frequency, cash advance frequency and, purchases installments frequency.

In advance frequency histogram shows that most customers do not like to pay in advance.

In the frequency of purchase installment, the histogram shows that most of the customers pay a monthly installment.

In purchase frequency, the histogram shows that most of the customers are just started using a credit card. whereas, most of the customers are regular customers.

In onoff purchases, the frequency histogram shows that most of the customers do not like purchasing in one go.

In purchases, the histogram shows that most of the customer's purchase amount to less than 5000.

In balances, the histogram shows that most of the customer's balance amount is in the range of 0 to 5000.

In [23]:
sns.set(rc={'figure.figsize':(25,20)})
ax1 = plt.subplot2grid((4,8), (0,1), colspan=2)
ax2 = plt.subplot2grid((4,8), (0,3), colspan=2)
ax3 = plt.subplot2grid((4,8), (1,1), colspan=2)
ax4 = plt.subplot2grid((4,8), (1,3), colspan=2)
ax5 = plt.subplot2grid((4,8), (2,1), colspan=2)
ax6 = plt.subplot2grid((4,8), (2,3), colspan=2)
sns.set_theme(palette="Accent")
sns.histplot(data = credit_data,x = "PURCHASES",ax=ax5, color='yellow')
sns.histplot(data = credit_data,x = "PURCHASES_FREQUENCY",ax=ax3, color='red')
sns.histplot(data = credit_data,x = "BALANCE",ax=ax6, color='green')
sns.histplot(data = credit_data,x = "ONEOFF_PURCHASES_FREQUENCY",ax=ax4, color='blue')
sns.histplot(data = credit_data,x = "CASH_ADVANCE_FREQUENCY",ax=ax1, color='darkblue')
sns.histplot(data = credit_data,x = "PURCHASES_INSTALLMENTS_FREQUENCY",ax=ax2, color='purple');

These scatter plot shows that most customers have a credit limit of up to 20000, and payment is near up to 10000. whereas, in some cases, customers have credit limits of up to 30000.

In [76]:
sns.scatterplot(x=credit_data["CREDIT_LIMIT"],y=credit_data["PAYMENTS"]);
sns.set(rc={'figure.figsize':(6,9)})

Using the px function, plot the scatter plot for purchases and purchase frequency and customer id plots with a different color.

In this way, select any customer id and show what the purchase amount and purchase frequency are.

In [25]:
px.scatter(data_frame = credit_data
           ,x = 'PURCHASES'
           ,y = 'PURCHASES_FREQUENCY'
           ,color = 'CUST_ID'
           )

Plot the correlation using a heatmap.

The histogram shows that only Purchases and Onoff purchase frequency are related. whereas, the other columns are not corelated in this histogram.

In [26]:
plt.figure(figsize=(13, 7))
sns.heatmap(credit_data.corr(), annot=True);

In the credit limit, the box plot shows that the median credit limit is around 3000. whereas, the third quartile is nearly 7000.

There are many outliers in this data up to 30000.

In [27]:
sns.boxplot(credit_data['CREDIT_LIMIT']);
sns.set(rc={'figure.figsize':(8,10)});

/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning:

Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.

Cluster Analysis¶

6)Step-by-step hierarchical clustering:¶

Take sample of data to find step-by-step hierarchical clustering

Make a new dataframe contain customer id, purchases, credit limit ,and balance.

In [28]:
from sklearn.cluster import AgglomerativeClustering
In [29]:
cl = credit_data[['CREDIT_LIMIT','PAYMENTS']]
In [30]:
cluster = AgglomerativeClustering(n_clusters =4, affinity = 'euclidean', linkage  = 'ward')
In [31]:
clc_pre = cluster.fit_predict(cl)
In [32]:
credit_data['cluster']=clc_pre
In [33]:
credit_data[['CREDIT_LIMIT','PURCHASES','cluster']].head()
Out[33]:
CREDIT_LIMIT PURCHASES cluster
0 1000.0 95.40 1
1 7000.0 0.00 0
2 7500.0 773.17 0
3 7500.0 1499.00 0
4 1200.0 16.00 1

Import AgglomerativeClustering from sklearn.cluster select two column credit_limit and payments for clustering then fit the cluster.

In [34]:
df1 = credit_data[credit_data.cluster == 0]
df2 = credit_data[credit_data.cluster == 1]
df3 = credit_data[credit_data.cluster == 2]
df4 = credit_data[credit_data.cluster == 3]
In [35]:
plt.scatter(df1['CREDIT_LIMIT'],df1['PAYMENTS'],color ='black')
plt.scatter(df2['CREDIT_LIMIT'],df2['PAYMENTS'],color ='orange')
plt.scatter(df3['CREDIT_LIMIT'],df3['PAYMENTS'],color ='purple')
plt.scatter(df4['CREDIT_LIMIT'],df4['PAYMENTS'],color ='blue');

7)Dendogram¶

for plot the dendogram import dendrogram,linkage and leaves_ list from the scipy cluster hierarchy. we take sample data to plot the dendogram to allocate objects to clusters in dendogram.

In [36]:
u_updated_credit_data = credit_data.drop(['CUST_ID','High_net_indi'],axis =1)
In [37]:
from scipy.cluster.hierarchy import dendrogram, linkage, leaves_list
Z = linkage(u_updated_credit_data.sample(80), 'average')


plt.figure(figsize=(25, 10))
plt.title('Hierarchical Clustering Dendrogram')
dendrogram(Z, leaf_font_size = 19.)
plt.show()

8)Unsupervised Clustering¶

step-by-step k-means¶

decide to use 3 centroid for the data and select random points.

In [38]:
centroid1 = np.array([8000,7000])
centroid2 = np.array([15000, 17000])
centroid3 = np.array([17000, 30000])
print(centroid1, centroid2, centroid3)

[8000 7000] [15000 17000] [17000 30000]

plot the scatter plot for the credit card limit and payments columns. and plot the 3 centroid in the scatter plot.

In [39]:
fig = plt.figure(figsize=(10,10))

plt.scatter(credit_data.CREDIT_LIMIT, credit_data.PAYMENTS, edgecolor = 'black', s = 90)
plt.title('USERS LIMITS & PAYMENTS DETAILS')
plt.xlabel('CREDIT_LIMIT')
plt.ylabel('PAYMENTS')
circle1 = plt.Circle(centroid1, radius=1000, edgecolor = 'black', fc='yellow')
circle2 = plt.Circle(centroid2, radius=1000, edgecolor = 'black', fc='red')
circle3 = plt.Circle(centroid3, radius=1000, edgecolor = 'black', fc='green')
plt.gca().add_patch(circle1)
plt.gca().add_patch(circle2)
plt.gca().add_patch(circle3)
Out[39]:
<matplotlib.patches.Circle at 0x7f607c525cd0>
In [40]:
credit_data.head()
Out[40]:
CUST_ID BALANCE BALANCE_FREQUENCY PURCHASES ONEOFF_PURCHASES INSTALLMENTS_PURCHASES CASH_ADVANCE PURCHASES_FREQUENCY ONEOFF_PURCHASES_FREQUENCY PURCHASES_INSTALLMENTS_FREQUENCY CASH_ADVANCE_FREQUENCY CASH_ADVANCE_TRX PURCHASES_TRX CREDIT_LIMIT PAYMENTS MINIMUM_PAYMENTS High_net_indi Discount cluster
0 C10001 40.900749 0.818182 95.40 0.00 95.4 0.000000 0.166667 0.000000 0.083333 0.000000 0 2 1000.0 201.802084 139.509787 Regular customer 0 1
1 C10002 3202.467416 0.909091 0.00 0.00 0.0 6442.945483 0.000000 0.000000 0.000000 0.250000 4 0 7000.0 4103.032597 1072.340217 Regular customer 0 0
2 C10003 2495.148862 1.000000 773.17 773.17 0.0 0.000000 1.000000 1.000000 0.000000 0.000000 0 12 7500.0 622.066742 627.284787 Regular customer 0 0
3 C10004 1666.670542 0.636364 1499.00 1499.00 0.0 205.788017 0.083333 0.083333 0.000000 0.083333 1 1 7500.0 0.000000 0.000000 Regular customer 0 0
4 C10005 817.714335 1.000000 16.00 16.00 0.0 0.000000 0.083333 0.083333 0.000000 0.000000 0 1 1200.0 678.334763 244.791237 Regular customer 0 1

Import make_blobs from sklearn.cluster.

make blobs with sample of 120 and with the centers of 4.

For finding kmenas import KMenas from sklearn.cluster.

In [41]:
 from sklearn.datasets import make_blobs
In [42]:
dataset = make_blobs(n_samples=120, centers=4,
                       cluster_std=0.40, random_state=0)
In [43]:
points = dataset[0]
In [43]:
In [44]:
from sklearn.cluster import KMeans
In [45]:
kmeans = KMeans(n_clusters=4)

remove the null values from the datsets.

In [46]:
np.isnan(credit_data.any()) #and gets False
np.isfinite(credit_data.all()) #and gets True
Out[46]:
CUST_ID                             True
BALANCE                             True
BALANCE_FREQUENCY                   True
PURCHASES                           True
ONEOFF_PURCHASES                    True
INSTALLMENTS_PURCHASES              True
CASH_ADVANCE                        True
PURCHASES_FREQUENCY                 True
ONEOFF_PURCHASES_FREQUENCY          True
PURCHASES_INSTALLMENTS_FREQUENCY    True
CASH_ADVANCE_FREQUENCY              True
CASH_ADVANCE_TRX                    True
PURCHASES_TRX                       True
CREDIT_LIMIT                        True
PAYMENTS                            True
MINIMUM_PAYMENTS                    True
High_net_indi                       True
Discount                            True
cluster                             True
dtype: bool

select the numeric columns from the data.

In [47]:
x = credit_data.iloc[ : , 1:13]
In [48]:
kmeans.fit(x)
Out[48]:
KMeans(n_clusters=4)
In [49]:
kmeans.labels_
Out[49]:
array([0, 1, 0, ..., 0, 0, 0], dtype=int32)

plot the KMeans classification using the scatter plot for the purchases and purchases frequency.

In [50]:
fig = plt.figure(figsize=(8, 13))
colormap = np.array(['green', 'red', 'blue','magenta','black'])
 
plt.scatter(x.PURCHASES_FREQUENCY, x.PURCHASES,c = colormap[kmeans.labels_], edgecolor = 'black', s=80)
plt.title('K Mean Classification')
plt.xlabel('PURCHASS')
plt.ylabel('PURCHASES_FREQUENCY')
Out[50]:
Text(0, 0.5, 'PURCHASES_FREQUENCY')

plot the kmeans four different clusters.

In [51]:
clusters = kmeans.cluster_centers_
In [52]:
print(clusters)

[[8.94422232e+02 8.55163052e-01 6.10121940e+02 3.09709112e+02
  3.00685549e+02 4.31797717e+02 4.81657254e-01 1.69810902e-01
  3.58144038e-01 9.78363640e-02 2.00318295e+00 1.09868530e+01]
 [5.17272254e+03 9.66444280e-01 5.81851909e+02 3.34496371e+02
  2.47415529e+02 4.78561083e+03 3.19561219e-01 1.45745998e-01
  2.25278073e-01 4.14781650e-01 1.21612903e+01 1.00519713e+01]
 [2.80368622e+03 9.77170735e-01 5.62444338e+03 3.73408263e+03
  1.89146826e+03 4.73842877e+02 9.07397795e-01 6.88976375e-01
  6.94633205e-01 6.61468259e-02 1.68771331e+00 6.53071672e+01]
 [5.60164463e+03 9.54132227e-01 2.83941623e+04 2.28584573e+04
  5.53570500e+03 1.01420640e+03 9.00757545e-01 8.43181864e-01
  6.95454500e-01 4.16666364e-02 1.86363636e+00 1.26090909e+02]]

Prediction¶

In [53]:
y_km = kmeans.fit_predict(x)
X = np.array(x)
In [54]:
fig = plt.figure(figsize=(8, 13))
plt.scatter(X[y_km == 0,0], X[y_km == 0,1], s=50, color='blue', label='Cluster0')
plt.scatter(X[y_km == 1,0], X[y_km == 1,1], s=50, color='purple', label='Cluster1')
plt.scatter(X[y_km == 2,0], X[y_km == 2,1], s=50, color='green', label='Cluster2')
plt.scatter(X[y_km == 3,0], X[y_km == 3,1], s=50, color='red', label='Cluster3')

plt.scatter(clusters[0][0], clusters[0][1], marker='*', s=200, color='black')
plt.scatter(clusters[1][0], clusters[1][1], marker='*', s=200, color='black')
plt.scatter(clusters[2][0], clusters[2][1], marker='*', s=200, color='black')
plt.scatter(clusters[3][0], clusters[3][1], marker='*', s=200, color='black')
plt.show()

9)sklearn.cluster¶

Import cluster and metrics from sklearn.

In [55]:
import sklearn.cluster as cluster
import sklearn.metrics as metrics

Import KNeighborsClassifier from sklearn.neighbors.

In [56]:
from sklearn.neighbors import KNeighborsClassifier

Find the kmeans of the clusters for 4 clusters. and fit the kmeans with purchases and purchses frequency.

In [57]:
kmeans = cluster.KMeans(n_clusters=4, init="k-means++").fit(X)
kmeans =kmeans.fit(x[['PURCHASES','PURCHASES_FREQUENCY']])

Find the sklearn cluster with the cluster center and plot the scatter plot for purchases installment frequency and cash advance frequency for the difference ranges of balances.

In [58]:
skcluster= kmeans.cluster_centers_
In [59]:
skcluster
Out[59]:
array([[3.79142365e+02, 4.08404888e-01],
       [8.47365443e+03, 9.26444251e-01],
       [2.83571611e+03, 8.64829872e-01],
       [2.86012077e+04, 9.15909045e-01]])
In [60]:
sns.scatterplot(x="PURCHASES_INSTALLMENTS_FREQUENCY", y="CASH_ADVANCE_FREQUENCY", hue = 'BALANCE',  data=x, palette='viridis');

10) KNN algorithm¶

import essential library such as matplotlib, KMeans and sklearn metrics.

In [61]:
import matplotlib.pyplot as plt
from matplotlib.colors import ListedColormap
from sklearn import datasets
from sklearn.cluster import KMeans
import sklearn.metrics as sm
 
import pandas as pd
import numpy as np
from pandas import DataFrame

# Set some pandas options
pd.set_option('display.notebook_repr_html', False)
pd.set_option('display.max_rows', 60)
pd.set_option('display.max_columns', 60)
pd.set_option('display.width', 1000)
 
%matplotlib inline

convert purchase frequency into the 0,1,2,3.

In [62]:
credit_data['convert_purchase'] = 0
for i in range(0,8950):
  if credit_data.iloc[i,7] >= 0.75:
    credit_data.iloc[i,18] = 3
  elif credit_data.iloc[i,7] >= 0.50:
    credit_data.iloc[i,18] = 2  
  elif credit_data.iloc[i,7] >= 0.25:
    credit_data.iloc[i,18] = 1   
  else:
    credit_data.iloc[i,18] = 0
In [63]:
credit_data.head()
Out[63]:
None
In [64]:
credit_data.dtypes
Out[64]:
CUST_ID                              object
BALANCE                             float64
BALANCE_FREQUENCY                   float64
PURCHASES                           float64
ONEOFF_PURCHASES                    float64
INSTALLMENTS_PURCHASES              float64
CASH_ADVANCE                        float64
PURCHASES_FREQUENCY                 float64
ONEOFF_PURCHASES_FREQUENCY          float64
PURCHASES_INSTALLMENTS_FREQUENCY    float64
CASH_ADVANCE_FREQUENCY              float64
CASH_ADVANCE_TRX                      int64
PURCHASES_TRX                         int64
CREDIT_LIMIT                        float64
PAYMENTS                            float64
MINIMUM_PAYMENTS                    float64
High_net_indi                        object
Discount                              int64
cluster                               int64
convert_purchase                      int64
dtype: object

Plot the

In [66]:
from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier(n_neighbors = 5, 
                           p = 2)# p=2 for euclidean distance
knn.fit(credit_data[["PURCHASES", "BALANCE"]], 
        credit_data.convert_purchase)
Out[66]:
KNeighborsClassifier()
In [68]:
credit_data['convert_purchase'] = 0
for i in range(0,8950):
  if credit_data.iloc[i,7] >= 0.75:
    credit_data.iloc[i,19] = 3
  elif credit_data.iloc[i,7] >= 0.50:
    credit_data.iloc[i,19] = 2  
  elif credit_data.iloc[i,7] >= 0.25:
    credit_data.iloc[i,19] = 1   
  else:
    credit_data.iloc[i,19] = 0
In [69]:
fig = plt.figure(figsize=(5,5))

<Figure size 360x360 with 0 Axes>
In [70]:
colormap = np.array(['black', 'yellow', 'red', 'blue', 'green'])
 
plt.scatter(credit_data.sample(80).CREDIT_LIMIT,credit_data.sample(80).PURCHASES, 
            c = colormap[credit_data.sample(80).convert_purchase], edgecolor = 'black', s = 120)
plt.title('Purchases:\nPurchases 1 - Yellow\nPurchases 2 - Red\nPurchases 3 - Blue\nPurchases 4 - green')
ax = fig.add_subplot(111)
ax.set_xlabel('Credit Limit')
ax.set_ylabel('Purchases')
Out[70]:
Text(0, 0.5, 'Purchases')
In [79]:
fig = plt.figure(figsize=(8,7))

<Figure size 576x504 with 0 Axes>
In [81]:
colormap = np.array(['black', 'yellow', 'red', 'blue', 'green'])
 
plt.scatter(credit_data.CREDIT_LIMIT,credit_data.PURCHASES, 
            c = colormap[credit_data.convert_purchase], edgecolor = 'black', s = 120)
plt.title('Purchases:\nPurchases 1 - Yellow\nPurchases 2 - Red\nPurchases 3 - Blue\nPurchases 4 - green')
ax = fig.add_subplot(111)
ax.set_xlabel('Credit Limit')
ax.set_ylabel('Purchases')

/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:6: MatplotlibDeprecationWarning:

Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.
Out[81]:
Text(0, 0.5, 'Purchases')
In [73]:
from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier(n_neighbors = 5, 
                           p = 2)# p=2 for euclidean distance
knn.fit(credit_data[["CREDIT_LIMIT", "PURCHASES"]], 
        credit_data.convert_purchase)
Out[73]:
KNeighborsClassifier()
In [74]:
def plotMesh():
    h = 100 
    # Create color maps
    cmap_light = ListedColormap(['#ffffb3', '#ff9999','#d6d6f5','#ccffdd'])
    colormap = np.array(['black', 'yellow', 'red', 'blue', 'green'])
     
    x_min, x_max = credit_data.CREDIT_LIMIT.min() - 1000, credit_data.CREDIT_LIMIT.max() + 1000
    y_min, y_max = credit_data.PURCHASES.min() - 1000, credit_data.PURCHASES.max() + 1000
    xx, yy = np.meshgrid(np.arange(x_min, x_max, h),np.arange(y_min, y_max, h))
    Z = knn.predict(np.c_[xx.ravel(), yy.ravel()])

    # Put the result into a color plot
    Z = Z.reshape(xx.shape)
    fig = plt.figure(figsize=(13,19))
    plt.pcolormesh(xx, yy, Z, cmap = cmap_light)
    # Plot also the training points
    plt.scatter(credit_data.CREDIT_LIMIT, credit_data.PURCHASES, c = colormap[credit_data.convert_purchase], edgecolor = 'black', 
                s = 120)
    plt.xlim(xx.min(), xx.max())
    plt.ylim(yy.min(), yy.max())
    plt.title("4-Class classification \n(k = %i)\n\
    Loan 1 - Yellow, Loan 2 - Red, Loan 3 - Blue, Loan 4 - green" % (5))
    ax = fig.add_subplot(111)
    ax.set_xlabel('Purchases')
    ax.set_ylabel('Balance')
plotMesh()

/usr/local/lib/python3.7/dist-packages/sklearn/base.py:451: UserWarning:

X does not have valid feature names, but KNeighborsClassifier was fitted with feature names

/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:23: MatplotlibDeprecationWarning:

Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.
In [ ]:

Report¶

credit card.jfif

This Report is created by Group of 9 analytical computing. For this project, Visa company is ours Stakeholder because they can easily understand the data based on our project and make important decisions easily.

1)Objective¶

  • The main objective of this research work is to analyze and represent the various factors that have contributed to the Evolution of credit card transaction over time in Graphical Format and perform a comparative study between various factors and read the credit card data file using read_csv command.

2) Data Exploration¶

  • Find the rows and column of the dataframe.Total number of rows is 8950,colunm =18.Find all statistical parameters such as mean, median, max, min, quartiles, and count for understanding data.After that,Check the types of the different parameters and found that every column has the proper type.

3) Data Cleaning¶

Find the data which has null value by using the isna function and found that minimum_payments column has 313 null value.Then, remove null value by passing 0 value.Drop the customer id and prc full payment column payment in the data.

4) Adaptation¶

Add a column about whether that customer is a regular customer or HNI.If the customer's bank balance is more than 2000, the customer is HNI otherwise, the customer is regular.the second added is about discount.Provide a discount to a customer when the purchase is more than 5000, so the discount is 500.whereas, if the purchase is more than 10000, so a discount is 1000.whereas, if the purchase is more than 20000 so, a discount is 2000.

5)Descriptive Visualization:¶

  • plot the histogram of the purchases, purchases frequency, balnace, oneoff purchases frequency, cash advance frequency and, purchases installments frequency.In advance frequency histogram shows that most customers do not like to pay in advance.In the frequency of purchase installment, the histogram shows that most of the customers pay a monthly installment.In purchase frequency, the histogram shows that most of the customers are just started using a credit card. whereas, most of the customers are regular customers.In onoff purchases, the frequency histogram shows that most of the customers do not like purchasing in one go.In purchases, the histogram shows that most of the customer's purchase amount to less than 5000.In balances, the histogram shows that most of the customer's balance amount is in the range of 0 to 5000.

  • These scatter plot shows that most customers have a credit limit of up to 20000, and payment is near up to 10000. whereas, in some cases, customers have credit limits of up to 30000.Using the px function, plot the scatter plot for purchases and purchase frequency and customer id plots with a different color. In this way, select any customer id and show what the purchase amount and purchase frequency are.Plot the correlation using a heatmap.The histogram shows that only Purchases and Onoff purchase frequency are related. whereas, the other columns are not corelated in this histogram.In the credit limit, the box plot shows that the median credit limit is around 3000. whereas, the third quartile is nearly 7000.There are many outliers in this data up to 30000.

In [85]:
sns.set(rc={'figure.figsize':(16,12)})
ax1 = plt.subplot2grid((4,8), (0,1), colspan=2)
ax2 = plt.subplot2grid((4,8), (0,3), colspan=2)
ax3 = plt.subplot2grid((4,8), (1,1), colspan=2)
ax4 = plt.subplot2grid((4,8), (1,3), colspan=2)
ax5 = plt.subplot2grid((4,8), (2,1), colspan=2)
ax6 = plt.subplot2grid((4,8), (2,3), colspan=2)
sns.set_theme(palette="Accent")
sns.histplot(data = credit_data,x = "PURCHASES",ax=ax5, color='yellow')
sns.histplot(data = credit_data,x = "PURCHASES_FREQUENCY",ax=ax3, color='red')
sns.histplot(data = credit_data,x = "BALANCE",ax=ax6, color='green')
sns.histplot(data = credit_data,x = "ONEOFF_PURCHASES_FREQUENCY",ax=ax4, color='blue')
sns.histplot(data = credit_data,x = "CASH_ADVANCE_FREQUENCY",ax=ax1, color='darkblue')
sns.histplot(data = credit_data,x = "PURCHASES_INSTALLMENTS_FREQUENCY",ax=ax2, color='purple');
In [89]:
px.scatter(data_frame = credit_data
           ,x = 'PURCHASES'
           ,y = 'PURCHASES_FREQUENCY'
           ,color = 'CUST_ID'
           )

Cluster Analysis¶

6)Step-by-step hierarchical clustering:¶

import agglomerativeclustering from sklearn.cluster.Then the select two column credit_limit and payments for clustring.then fit to the cluster.Build four cluster df1, df2,df3, df4.Lastly, plot the scatter plot for the four different clusters and it is represented by different colors.

7)Dendogram¶

for plot the dendogram import dendrogram,linkage and leaves_ list from the scipy cluster hierarchy.We take sample data to plot the dendogram to allocate objects to clusters in dendogram.

In [82]:
from scipy.cluster.hierarchy import dendrogram, linkage, leaves_list
Z = linkage(u_updated_credit_data.sample(80), 'average')


plt.figure(figsize=(25, 10))
plt.title('Hierarchical Clustering Dendrogram')
dendrogram(Z, leaf_font_size = 19.)
plt.show()

8)Unsupervised Clustering¶

decide to use 3 centroid for the data and select random points.plot the scatter plot for the credit card limit and payments columns.and plot the 3 centroid in the scatter plot. import make_blobs from sklearn.cluster.make blobs with sample of 120 and with the centers of 4.For finding kmenas import KMenas from sklearn.cluster.select the numeric columns from the data.plot the KMeans classification using the scatter plot for the purchases and purchases frequency.print four different clusters of the KMeans.

In [91]:
fig = plt.figure(figsize=(7,6))

plt.scatter(credit_data.CREDIT_LIMIT, credit_data.PAYMENTS, edgecolor = 'black', s = 90)
plt.title('USERS LIMITS & PAYMENTS DETAILS')
plt.xlabel('CREDIT_LIMIT')
plt.ylabel('PAYMENTS')
circle1 = plt.Circle(centroid1, radius=1000, edgecolor = 'black', fc='yellow')
circle2 = plt.Circle(centroid2, radius=1000, edgecolor = 'black', fc='red')
circle3 = plt.Circle(centroid3, radius=1000, edgecolor = 'black', fc='blue')
plt.gca().add_patch(circle1)
plt.gca().add_patch(circle2)
plt.gca().add_patch(circle3)
Out[91]:
<matplotlib.patches.Circle at 0x7f607b484f90>

9)sklearn.cluster¶

import cluster and metrics from sklearn.import KNeighborsClassifier from sklearn.neighbors.find the kmeans of the clusters for 4 clusters.fit the kmeans with purchases and purchses frequency.Find the sklearn cluster with the cluster center and plot the scatter plot for purchases installment frequency and cash advance frequency for the difference ranges of balances.

In [92]:
sns.scatterplot(x="PURCHASES_INSTALLMENTS_FREQUENCY", y="CASH_ADVANCE_FREQUENCY", hue = 'BALANCE',  data=x, palette='viridis');

10) KNN algorithm¶

import essential library such as matplotlib, KMeans and sklearn metrics.convert purchase frequency into the 0,1,2,3 for plotting knn plot.Plot the scatter plot of the credit limit and purchases.

In [93]:
colormap = np.array(['black', 'yellow', 'red', 'blue', 'green'])
 
plt.scatter(credit_data.sample(80).CREDIT_LIMIT,credit_data.sample(80).PURCHASES, 
            c = colormap[credit_data.sample(80).convert_purchase], edgecolor = 'black', s = 120)
plt.title('Purchases:\nPurchases 1 - Yellow\nPurchases 2 - Red\nPurchases 3 - Blue\nPurchases 4 - green')
ax = fig.add_subplot(111)
ax.set_xlabel('Credit Limit')
ax.set_ylabel('Purchases')

/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:6: MatplotlibDeprecationWarning:

Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.
Out[93]:
Text(20.200000000000003, 0.5, 'Purchases')

Originality¶

Originality of the project - This dataset was not created by us rather it was imported from Kaggle "https://www.kaggle.com/datasets/arjunbhasin2013/ccdata?resource=download" . We did not refer or copy anyones work from the internet. But, there might be some scenarios where we could have used the same variables to predict the similar output as others on the web and also could have used the same visualization to present our graphs. All the notebooks that we came across were coded in python language hence different approach to dataset as the library used also differs to some extent. Although we reached to stack overflow and other multiple website for reference, this work solely belongs to us.

In [ ]: